Analysis of US Delegation’s Climate Numbers

By Claudia Tebaldi and Eric Larson

Climate Central scientists analyze the claims made at COP15 by the US Climate Change Special Envoy Todd Stern. For more information on this subject, please see Andrew Freedman's Blog: US Delegation’s Carbon Figures Check Out, But Mask Divides.

Measures Discussed by US Climate Change Special Envoy Todd Stern
#1: Total emissions relative to the 2005 level of emissions.
#2: Total emissions relative to the 1990 level of emissions.
#3: Total emissions reductions relative to a “business as usual” future
#4: Reductions in average per-capita emissions relative to 1990 levels
#5: Reductions in average emissions per $ of gross domestic product (so-called “carbon intensity”)
#6: Reductions in atmospheric concentrations of greenhouse gases.

Claims #1 and #2 by Stern

US emissions reduction targets, relative to actual 2005 emissions, are 17% lower by 2020 and 42% lower by 2030. The latter number corresponds to a 33% reduction relative to actual 1990 emissions.The 2020 US target (17% reduction relative to 2005) is higher than the corresponding EU target (13% relative to 2005 emissions), though lower than the EU target when 1990 is the baseline.

Climate Central says:

Stern's claims are correct.  It is also true that if a 1990 baseline is used, the US's proposed reduction is 4% by 2020, whereas the EU's proposed reduction is 20% by 2020.  Now for the calculations:

> Greenhouse gas emissions in 1990 were 6.148 billion metric tonnes CO2 equivalent (CO2e).
> Greenhouse gas emissions in 2005 were 7.130 billion metric tonnes CO2e.
   Source: EPA US Greenhouse Gas Inventory

So, if US emissions in 2030 are going to be 42% less than in 2005, the emissions in 2030 will be 7.130 * (1 - 0.42) = 4.135 billion t CO2e.

As a fraction of the 1990 emission level, this would be 4.135 / 6.148 = 0.67. Said another way, emissions in 2030 would be 67% of emissions in 1990, or the reduction in emissions would be 33% compared to the 1990 level (confirming Stern’s claim). 

When the same calculation is made for the US target for 2020, it computes to a 4% reduction compared to the US's 1990 level.  The EU target for 2020 is a 20% reduction compared to the EU's 1990 level.

To determine EU emissions in 1990 and 2005, we added up greenhouse gas emissions for all 25 countries currently in the European Union (excluding Malta/Cyprus for which numbers are not available). Annual emissions by country are compiled by the UNFCCC. This gives the result that emissions for the EU in 1990 and in 2005 were 5.742 billion t CO2e and 5.108 billion t CO2e, respectively (we used the values in Table 4 of the UNFCC report, excluding sources/sinks from Land Use, Land Use Change and Forestry). The EU has reduced its emissions between 1990 and 2005. Using the same calculation as above for the US, one can conclude that an EU emissions reduction by 2020 that is 13% of the 2005 emissions level would be 20% of the 1990 emissions level (confirming Stern’s claim).

Note that because EU emissions were lower in 2005 than in 1990 (as a result of efforts in the EU to comply with its Kyoto Protocol commitments), the percentage reduction achieved in 2020 is smaller when measured relative to 2005 than relative to 1990. On the other hand, because emissions increased from 1990 to 2005 in the US, the exact opposite is true: the percentage reduction in 2020 for the US is larger when measured relative to 2005 than when measured relative to 1990.

Claim #3 by Stern

If compared to business-as-usual (BAU) the reduction in 2020 would be 17% for the US, 12% for the EU, 10% for Japan, and 20% for Australia.

Climate Central says:

This claim cannot be checked without knowing what the BAU emissions are for 2020 in each of these countries. In his comments Stern did not give an indication of BAU assumptions for 2020.

Claim #4 by Stern

When measured on a per-capita basis, the US emission reduction target for 2020 would represent a 29% reduction relative to 1990 per-capita emissions. This compares with a 25% reduction for the same metric in the case of the EU.

Climate Central says:

Checking this claim requires population estimates for the US for 1990 and 2020.

> US population in 1990 was 250.1 million. (Source: US Census)
> US population projected for 2020 is 341.4 million (Source: US Census)

US emissions in 1990 were 6.148 billion t CO2e (see Claim #1 above), so per-capita emissions in 1990 were 6.148 billion / 250.1 million = 24.6 t CO2e per capita. US emissions in 2020 under the proposed reduction target would be 17% less than 2005 emissions. Emissions in 2005 were 7.130 billion t CO2e (see Claim #1 above), so 2020 emissions would be 7.130 x (1 – 0.17) = 5.918 billion t CO2e. Thus, per-capita emissions in 2020 would be 5.918 billion / 341.4 million = 17.3 t CO2e per capita. So per-capita emission in 2020 would be (24.6 – 17.3) / 24.6 = 0.29, or 29% less than in 1990 (as claimed by Stern).

For the EU, our computations use population projections by country provided by the UN “World Population Prospects: The 2006 Revision Population Database” (UN Department of Economic and Social Affairs). We sum the population size of the same 25 EU member countries that entered our computations of total emissions (see Claim #1 above), obtaining 470.5 million as the EU population in 1990 and 503.9 million as the projections for 2020.  (There were fewer EU members in 1990 than today, but we use the same 25 for all computations for consistency.)

Using the EU emission estimates of 5.742 billion t CO2e for 1990 and of 4.443 billion t CO2e for 2020, we compute per capita emissions in 1990 of 5.742 billion/470.5 million = 12.2 t CO2e and 4.443 billion/503.9 million = 8.8 t CO2e. Thus the projected EU cut in per capita emissions in 2020 with respect to the 1990 value would be (12.2-8.8)/12.2= 0.28, or 28% less than 1990, confirming once again Stern’s numbers.  The US per capita target compares favorably against the EU target, even with a 1990 baseline, because the population of the US is projected to grow by a much larger percentage.

Claim #5 by Stern

The reduction by 2020 in carbon intensity, i.e., carbon emissions per unit of GDP, is comparable for the US to targets proposed by the EU and other developed countries.

Climate Central says:

The calculation requires estimates of GDP. For 1990 and 2005, US GDP was $5.8 and $12.4 trillion (in contemporary dollars), or $7.1 and $11.0 trillion, respectively, in constant year-2000 dollars. (Source: US Census)

From analysis of Claim #1 (above), the carbon intensities per unit of constant-dollar GDP were

> 1990: 6.148 billion t CO2e / $7.1 trillion = 0.86 t CO2e per 1000$
> 2005: 7.130 billion t CO2e / $11 trillion = 0.65 t CO2e per 1000$

If we assume a plausible average GDP growth rate of 3%/yr from 2005 to 2020, then GDP in 2020 would be 11 x (1+.03)15 = $17.1 trillion. Since 2020 emissions are 5.918 billion t CO2e for the US target indicated by Stern (see Claim #4), carbon intensity in 2020 would be 5.918 / 17.1 = 0.35 t CO2e per 1000$. This represents a 46% reduction in carbon intensity relative to 2005. Shortly before the start of COP15, China publicly stated their aim to reduce carbon intensity 45% to 50% by 2020.

Separately, using historical (2000) and projected (2010, 2020) GDP from this source, we can make a comparative estimate of US and EU carbon intensities.  We use the average of 2000 and 2010 to get a consistent estimate for 2005, resulting in the following numbers:

  US EU  
2005 $11.5 trillion $11.2 trillion
2020 $16.4 trillion $16.9 trillion


  US EU  
2005 0.62 t CO2e / 1000$ 0.46 t CO2e / 1000$
2020 0.36 t CO2e / 1000$ 0.26 t CO2e / 1000$


With these estimates, the reduction in carbon intensity in 2020 relative to 2005 is 42% for US and 43% for EU, confirming Sterns claim.

Claim #6 by Stern

In terms of targets for atmosphere concentration, the US proposal would correspond to a reduction 3 times as large as proposed by other industrialized countries, even after accounting for different starting points.

Climate Central says:

Climate Central is unable to assess this claim, because the different starting points are not specified.